Mailinglist Archive: opensuse-factory (715 mails)

[Richard Guenther <rguenther@xxxxxxx>]

On Thu, 15 Mar 2012, Dominique Leuenberger a.k.a DimStar wrote:

> Quoting "Dominique Leuenberger a.k.a DimStar" <DimStar@xxxxxxxxxxxx>:
>
> > On Stephan's great recommendation I went through the list of packages
> > failing in Gcc47 Staging area that were not currently listed as 'failed' in
> > openSUSE:Factory. Assuming that this is the list of packages to work on, I
> > branched them into home:dimstar:gcc47
> >
> > This list is now 'shorter' (currently only 77 packages, but Gcc Staging
> > still had > 3000 packages 'scheduled' to let fail). IT's good to have a base
> > to work on at least.
> >
> > Anybody else volunteering to work on gcc 4.7 issues, I recommend to work
> > with me on this 'single list' instead of redoing the work in several places.
> > I have no problem of granting anybody access to this repo, so we don't have
> > to go through SRs to speed up things.
> >
> > From this repo, all packages will need to be SRed to their respective devel
> > projects. This should keep us on track I hope.
> >
> > Looking forward to all your help! I will certainly not touch all of those
> > packages!
>
> Hi everybody!
>
> So, the gcc 4.7 testing does have some progress. So far there have been 21 SRs
> with fixes done from home:dimstar:gcc47 to their respective devel projects.
>
> This is quite good for the short time... but there is still a lot to do.
>
> At this point a big 'Thank you' to the helpers taking care of fixing those
> packages and also the reviewers spotting last-minute bugs introduced in
> patches... your work is very valuable in this move to gcc 4.7.
>
> About home:dimstar:gcc47:
> This project has been setup to link all the packages that currently fail in
> openSUSE:Factory:Staging:Gcc47.. But: it does not link the version that is
> currently in Factory, but the package from the respective devel-repo.
>
> Like this we were already able to identify a few packages that would fail in
> Factory with gcc 4.7, but have been updated to later version and do build in
> my branch! That's great... those packages just would need to be forwarded
> (examples packages of this kind are libwebkit/libwebkit3 and boost.)
>
> Unfortunately, with all those good news, I'm afraid we're also hit by a bug in
> gcc :P
>
> Especially the build of llvm aborts with a very strange error of undefined
> symbols, done in the gcc include tree:
>
> make[2]: *** [lib/Support/CMakeFiles/LLVMSupport.dir/APInt.cpp.o] Error 1
> In file included from
> /home/abuild/rpmbuild/BUILD/llvm-3.0/lib/Support/Allocator.cpp:16:
> In file included from
> /home/abuild/rpmbuild/BUILD/llvm-3.0/include/llvm/Support/Recycler.h:18:
> In file included from
> /home/abuild/rpmbuild/BUILD/llvm-3.0/include/llvm/ADT/ilist.h:44:
> In file included from
> /usr/lib/gcc/i586-suse-linux/4.7/../../../../include/c++/4.7/iterator:63:
> In file included from
> /usr/lib/gcc/i586-suse-linux/4.7/../../../../include/c++/4.7/ostream:39:
> In file included from
> /usr/lib/gcc/i586-suse-linux/4.7/../../../../include/c++/4.7/ios:42:
> In file included from
> /usr/lib/gcc/i586-suse-linux/4.7/../../../../include/c++/4.7/bits/ios_base.h:40:
> /usr/lib/gcc/i586-suse-linux/4.7/../../../../include/c++/4.7/ext/atomicity.h:48:45:
> error: use of undeclared identifier '__ATOMIC_ACQ_REL'
>  { return __atomic_fetch_add(__mem, __val, __ATOMIC_ACQ_REL); }
>                                            ^
> /usr/lib/gcc/i586-suse-linux/4.7/../../../../include/c++/4.7/ext/atomicity.h:52:38:
> error: use of undeclared identifier '__ATOMIC_ACQ_REL'
>  { __atomic_fetch_add(__mem, __val, __ATOMIC_ACQ_REL); }
>                                     ^
> 2 errors generated.
> make[2]: *** [lib/Support/CMakeFiles/LLVMSupport.dir/APFloat.cpp.o] Error 1
> 2 errors generated.
> make[2]: *** [lib/Support/CMakeFiles/LLVMSupport.dir/Allocator.cpp.o] Error 1
> make[1]: *** [lib/Support/CMakeFiles/LLVMSupport.dir/all] Error 2
> make: *** [all] Error 2
> error: Bad exit status from /var/tmp/rpm-tmp.MmKjDo (%build
>
>
> => I really doubt that this is something that should be fixed in llvm, but the
> gcc include file might be missing something more... I will glady hear
> different opinions or even better, see patches :P

You know the drill ... open a bugzilla and attach preprocessed source
of the file that exhibits this error.

Richard.

-- 
Richard Guenther <rguenther@xxxxxxx>
SUSE / SUSE Labs
SUSE LINUX Products GmbH - Nuernberg - AG Nuernberg - HRB 16746
GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer